Intelligibility as a linear combination of dimensions in dysarthric speech

https://doi.org/10.1016/S0021-9924(02)00065-5Get rights and content

Abstract

Speech samples of 79 dysarthric patients (native speakers of English) were judged by two experienced judges by means of a perceptual rating scale covering the four main dimensions of speech production: voice quality, articulation, nasality and prosody as well as overall intelligibility. In order to determine the relative influence of the four basic dimensions on intelligibility in this study group, a multiple regression model was applied. This model shows that intelligibility can be expressed as a linear combination of weighted perceptual dimensions (R2=0.89). This expression reveals the relative importance of the individual dimensions on overall intelligibility. Articulation is shown to be the strongest contributor to intelligibility. A group of 16 dysarthric patients (native speakers of Dutch) was used as a control group in order to evaluate the validity of the expression. Judged and calculated ratings (within the 95% prediction intervals) were in agreement for 75% of the patients. These findings may contribute in the construction of a better balanced measure of functional intelligibility.

Learning outcomes

  • 1.

    Intelligibility can be expressed by a linear combination of weighted single dimensions indicating the relative impact of each of them.

  • 2.

    The impact of articulation on intelligibility is dominant but inclusion of the dimensions “nasality”, “voice” and “prosody” results in a more balanced estimation of intelligibility.

Introduction

Generally, the assessment of dysarthria relies on a combination of perceptual and instrumental analysis procedures, including acoustic analyses. Most of the perceptual scales are based on the studies by Darley, Aronson, and Brown (1975), using a number of speech dimensions to be rated on a severity rating scale. In the last decades, there has been an increasing interest in the measurement of intelligibility as a part of dysarthria assessment (Enderby, 1983, Gentil, 1992; Kent, Weismer, Kent, & Rosenbek, 1989; Kent et al., 1990; Yorkston & Beukelman, 1981; Ziegler, 1992).

Intelligibility can be defined as what is understood by the listeners of the phonetic realization of speech (Yorkston, Strand, & Kennedy, 1996). Comprehensibility on the other hand also includes the pragmatic and semantic aspects of language.

Intelligibility is a matter of major concern in patients with problems such as dysarthria because it is the correlate of a general communicative impairment (Weismer & Martin, 1992). Intelligibility scores allow one to determine the impact of a disorder on the communicative functioning of the speaker. Furthermore, it is an interesting tool to monitor the evolution of speech production and to determine an overall index of participation in spoken exchange of information (Ansel & Kent, 1992). Sometimes intelligibility is confused with articulation and phonology (Kent, Miolo, & Bloedel, 1994). In fact, it is the product of a series of interactive processes as phonation, articulation, resonance and prosody.

Most systems for the perceptual assessment are developed for English (Darley et al., 1975, Enderby, 1983) for French (Gentil, 1992) and German (Ziegler, Hartmann, & Wiesner, 1992). These tests, however, do not provide more than an estimate of severity (Weismer & Martin, 1992). Other authors focused on specific instruments to estimate intelligibility using lists of words or sentences (Platt, Andrews, Young, & Quinn, 1980; Tikofsky, 1970; Tikofsky & Tikofsky, 1964; Yorkston & Beukelman, 1981). Some systems make use of phonetic contrasts that are sensitive to dysarthric impairment and contribute significantly to intelligibility (Kent et al., 1989; Whitehill & Ciocca, 2000). This approach is limited to the articulation and word levels. The investigation of connected speech intelligibility in dysarthric populations is relatively limited because additional factors may contribute variability to the estimate of speech intelligibility. However, connected speech intelligibility might have more external validity, because the evaluation level is closer to the functional level of communication (Weismer, Jeng, Laures, Kent, & Kent, 2001).

Since speech is the result of a complex process, intelligibility measures preferably include aspects of all speech dimensions. Theodoros, Murdoch, and Chenerey (1994) found that in a group of head injured patients with overall intelligibility problems, patients also showed more than 90% of occurrence of the dimensions related to nasality and mixed nasality, resonance, articulation (consonant precision), prosody and voice (pitch variation) (Chenerey, 1998).

A recent review of acoustic studies of dysarthric speech by Kent, Kent, Vorperian, and Duffy (1999) showed that although a number of these studies on dysarthria have been published, the progress is slow. This is attributed to modest research efforts in this domain, the difficulty of acoustic analysis for speakers who may have phonatory, resonatory and articulatory disorders and the rather few published examples of broadly directed acoustic analyses of dysarthria.

Literature review reveals that there are few studies that take into account the dimensions of voice quality, nasality, articulation and prosody for the assessment of intelligibility. Kent et al. (1990) showed that most disrupted phonetic features of speech in patients with amyotrophic lateral sclerosis involved phonatory function and velopharyngeal valving next to lingual movements for the production of vowels and consonants. Their findings were based on a word intelligibility test. This research focuses on intelligibility and the relative impact of the main dimensions contributing to it. It is unclear to what extent each of these dimensions affect overall intelligibility of connected speech as judged by listeners. To explore this problem, the authors investigated how single dimensions are related to intelligibility, how they contribute to overall intelligibility and if it is possible to express intelligibility by a linear combination of single dimensions (voice, articulation, nasality and prosody).

The results of this research may contribute to a better understanding of the concept of “functional intelligibility” and the development of improved instruments for the objective assessment of dysarthric’s functional intelligibility.

Section snippets

Methods and materials

Speech samples of 79 dysarthric patients (native speakers of English) were randomly selected from the original Aronson’s (1993) tape recordings of several types of dysarthria. To make intelligibility judgments, the grandfather passage or the spontaneous speech passage of each patient was selected. The samples were judged by two experienced listeners familiar with dysarthric patients for more than 15 years. Both judges were not native speakers of English, but were extensively trained in the

Results

The correlation between the four dimensions and the overall judged intelligibility was calculated. The strongest correlation was found between articulation and intelligibility (0.82). For the other dimensions, the correlation seems to be less strong: 0.46 for voice quality, 0.32 for nasality and 0.55 for prosody.

Next to the ratings of the four dimensions, the judges indicated the most dominant dimension affecting intelligibility for each patient. Articulation (47%) is considered as the most

Discussion

The main purpose of the study was to explore the relative impact of voice quality, nasality, articulation and prosody on overall intelligibility in dysarthric patients.

In a first analysis the judgments of overall intelligibility and each of the dimensions were compared. Articulation and prosody show the strongest correlation with intelligibility while nasality the lowest. The judges impression regarding the impact of these dimensions on intelligibility reveals similar findings for articulation

Acknowledgements

This research was supported by a grant from the Antwerp University (UIA) for Dr. M. E. Hernández-Dı́az Huici.

References (22)

  • M.S. De Bodt et al.

    Test–retest study of the GRBAD scale: Influence of experience and professional background on perceptual rating of voice quality

    Journal of Voice

    (1997)
  • R.D. Kent et al.

    Acoustic studies of dysarthric speech: Methods, progress, and potential

    Journal on Communication Disorders

    (1999)
  • B.M. Ansel et al.

    Acoustic–phonetic contrasts and intelligibility in the dysarthria associated with mixed cerebral palsy

    Journal of Speech and Hearing Research

    (1992)
  • Aronson, A. E. (1993). Dysarthria: Differential diagnosis (audio tape). Rochester, MN: Mentor...
  • Chenerey, H. J. (1998). Perceptual analysis of dysarthric speech. In B. E. Murdoch (Ed.), Dysarthria: A physiological...
  • Darley, F. L., Aronson, A. E., & Brown, J. R. (1975). Motor speech disorders. Philadelphia: W.B....
  • Enderby, P. (1983) Frenchay dysarthria assessment. San Diego: College-Hill...
  • M. Gentil

    Phonetic intelligibility testing in dysarthria for the use of French language clinicians

    Clinical Linguistics and Phonetics

    (1992)
  • Hirano, M. (1981). Clinical examination of voice (pp. 81–84). New York:...
  • R.D. Kent et al.

    Impairment of speech intelligibility in men with amyotrophic lateral sclerosis

    Journal of Speech and Hearing Disorders

    (1990)
  • Kent, R. D., Miolo, G., & Bloedel, S. (1994). The intelligibility of children’s speech: A review of evaluation...
  • Cited by (147)

    • Are Acoustic Markers of Voice and Speech Signals Affected by Nose-and-Mouth-Covering Respiratory Protective Masks?

      2023, Journal of Voice
      Citation Excerpt :

      However, there are many other RPMs available (eg, cloth masks with/without air filtering pads), and it would be interesting to expand this study with other masks. Fourth, we analyzed RPM influence on 26 acoustic markers that are considered relevant to voice (eg, fO and voice quality), speech (eg, vowel differentiation and fricative distinction), and consequently also intelligibility.29 It would also be interesting to evaluate how listeners react to speakers wearing various nose-and-mouth-covering RPMs, and how aspects like voice quality, articulatory precision, nasality, speech intelligibility, speech acceptability, etc. are perceived in these sound-filtered speaker conditions.

    • A cross-linguistic perspective to classification of healthiness of speech in Parkinson's disease

      2022, Journal of Neurolinguistics
      Citation Excerpt :

      Thirdly, we hypothesized that, in line with observations by Näsström and Schalling (2020) and Hartelius et al. (2003), listeners with SLT experience would rely primarily on voice quality and phonation features when classifying dysarthric speech. We also expected that listeners with no training but with the similar language background would rely on the mix of phonation, articulation and prosody related features (DeBodt, Huici, & Van DeHeyning, 2002; Hartelius et al., 2003), while listeners unfamiliar with the speakers' language would rely more on universal components of prosodic insufficiencies such as monopitch (Whitehill, 2010; Pinto et al., 2017). We conducted an experiment to investigate which cues in speech lead listeners with different language backgrounds and expertise to classify healthiness of speech.

    View all citing articles on Scopus
    View full text